ISSN 1392–1320 MATERIALS SCIENCE (MEDŽIAGOTYRA). Vol. 12, No. 4. 2006
Carbonation of Alumina Cement-bonded Conventional
Refractory Castable in Fireplace
Ina PUNDIEN
Ė
∗, Stasys GOBERIS, Valentin ANTONOVI
Č
,
Rimvydas STONYS, Algimantas ŠPOKAUSKAS
Vilnius Gediminas Technical University of Thermal Insulation, Linkmenų 28, LT-08217 Vilnius, Lithuania
Received 31 August 2006; accepted 12 October 2006
It was determined by verification by in situ analysis that combustion products of firewood in fireplaces cause carbonation of alumina cement-bonded conventional refractory castable and accelerate its destruction. The carbonation of hydrated cement (Gorkal 40) paste was modeled in a laboratory conditions: having been fired at various temperatures (300 ºC – 950 ºC) the specimens were curing in humid conditions for two days. It was determined by X-ray analysis that the hydrates CAH10, C2AH8 and C3AH6 change into a mineral C12A7 after the cement paste are fired at 300 ºC – 800 ºC temperatures. The amount of the mineral depends on the firing temperature. In humid conditions C12A7 hydrates and carbonizes, consequently calcium carboaluminate hydrate forms. In the composition of cement and microsilica (9 %) carbonation of cement hydrates noticeably slow down at the same conditions. Carbonized minerals are not identified when the specimens are fired at 500 ºC –850 ºC temperatures.
Keywords: alumina cement-bonded castable, carbonation, X-ray analysis, microsilica.
INTRODUCTION
∗Domestic boilers, furnaces and fireplaces that are heated by biofuel (fire-wood, wood waste, peat) are widely used in Lithuania. The observations show, that the lining of such furnaces, which is made using conventional alumina cement-bonded castable, has not long durability. Having been exploited for some time, it destructs. In cases when gas or fuel oil is used to heat up furnaces, the exploitation period of analogical linings is 2 – 3 times longer than in cases when biofuel is used.
One of the reasons that cause this destruction might be carbonation of alumina cement hydration phases. It is known that the main hydration phases forms during the reaction between alumina cement and the water are as follow: CAH10 (forms at the temperature <21 ºC), C2AH8
and AH3 (21 – 35 ºC), C3AH6 and AH3 (>35 ºC) [1].
CAH10 and C2AH8 are meta-stable therefore depending on
temperature and other factors that change into C3AH6 and
AH3. It is indicated [2] that refractory castable start to
deteriorate by peeling-off of material due to carbonation if it is not exploited, i.e. not fired for a long time after its production. The practical observations [2] of refractory castable reveal the following tendencies:
1. Carbonation occurs easily in lightweight castables, but hardly at all in dense castables;
2. Carbonation occurs easily in not dried products and hardly at all in dried castables;
3. Carbonation occurs easily in an environment with high temperature and humidity;
4. When not dried products are exposed to rain, the carbonation rate is highly accelerated.
The carbonation of alumina cement’s phases is thought to occur by the following reactions [3]:
CAH10+CO2+xH2O → CaCO3+Al2O3⋅yH2O+(10+x-y)H2O; (1)
∗Corresponding author. Tel.: +370-5-2752231; fax.: +370-5-2731230.
E-mail address: futer@centras.lt (I. Pundienė)
C2AH8+2CO2+xH2O → 2CaCO3+Al2O3⋅yH2O+(8+x-y)H2O; (2) C3AH6+3CO2+xH2O → 3CaCO3+Al2O3⋅yH2O+(6+x-y)H2O. (3)
In various works [4 – 6] the influence of curing temperature (5 ºC – 60 ºC) on the carbonation of hydrates was analyzed. It is being stated that the rate of carbonation is higher at 20 ºC temperature when not stable CAH10 is in
composition. However, the carbonation of C2AH8 is
considerably higher when humidity is increased [2]. C3AH6
distinguishes as having the lowest rate of carbonation. It is known that under the influence of temperature hexagonal hydrates (CAH10 and C2AH8) converse to cubic
hydrates (C3AH6). For this reason the porosity of cement
paste increases and the compressive strength decreases. The literature [6 – 8] says that in most cases newly formed carbonates increase the compressive strength of such compositions. It is pointed out that a considerable amount of carbonates forms due to the fact that CO2 may penetrate
the porous structure of the material much easier.
In certain conditions carbonation causes an absolute destruction of the material. It was noticed [2] that the specimen of castable partially immersed into the water totally destructs due to strong carbonation in the area where it contacts with the water. Carbonation also causes a large-scale destruction of alumina cement materials be-cause Na+, K+ ions participate in alkaline hydrolysis [7-9].
Approximately 15 % of CO2 and 25 % of H2O form in
combustion products of firewood, therefore, it might be forecasted that a repeated hydration of cement’s minerals may occur in such conditions. Thus, due to the influence of CO2, H2O, KOH and NaOH, which are in combustion
products, the carbonation may start.
In order to determine the reasons of destruction that take place in the refractory products, which are made using alumina cement-bonded castable and exploited in furnaces and fireplaces, which are heated by burning wood waste, verification in site and laboratory analysis were carried out.
EXPERIMENTAL
Experimental work was performed using alumina cement “Gorkal 40”, which is manufactured in Poland. The content of Al2O3 in cement is not less than 40 %. The main
mineral phases are: CA, the ferrite phase, C12A7, C2AS;
bulk density is 1160 kg/m3, specific surface 0.310 m2/g –
0.380 m2/g, refractoriness – not less than 1280 °C.
Microsilica from “Elkem ASA (Norway), SiO2 makes
up not less then 98 % of its composition, bulk density 410 kg/m3, specific surface 15 m2/g, L.O.I ~2 % at the
temperature of 750 °C.
In order to determine the influence of temperature and water vapors upon the process of castable’s destruction, X-ray analysis of cement paste (composition A, alumina cement with W/C ratio = 1) and cement paste with the ad-ditive of 9 % of microsilica (composition B) was carried out. There were two stages of the specimens’ tests: first, the specimens that had cure 24 hours in normal condition and had been fired at various temperatures that varied from 300 °C up to 950 °C, were analyzed; second, the same specimens after firing had been curing in humid conditions (above water) for two days and only then their roentgeno-grams were recorded. The quantitative changes of separate minerals were estimated according to the intensity of diffraction peaks.
The X-ray analysis was carried out by X-ray-diffractometer DRON-2 with a cooper anticathode and nickel filter, the voltage of anode 30 kV, current of anode 8 mA, slits of the goniometer – 0.5; 1.0; 0.5 mm, the speed of recording of X-ray-diffraction patterns – 600 mm/min. The phase composition was defined using the ASTM card index of data.
In this work a conventional chamotte concrete of the following composition was used (mass %): chamotte 0 mm – 1 mm – 30; chamotte 1 mm – 5 mm – 30; disperse chamotte – 15; alumina cement – 25; water – 14.5.
Castable cubes of the size (70×70×70) mm were formed. After 3 days of normal curing, the samples were dried at a temperature of 105 °C ±5 °C for 48 hours. Next, they were kept for three hours under the investigated temperatures 300 °C, 600 °C and 800 °C respectively in an electronic controller furnace and cooled. Then the hardness of castable samples and cold compressive strength were determined. The hardness of the castable was estimated using Rockwell apparatus TP 5043 according to GOST 24622-81.
RESULTS AND DISCUSSION
Practical observation showed, that alumina cement-bonded castable blocks of boilers that are heated by biofuel disintegrate and fracture (Fig. 1).
In order to determine the reasons of destruction of conventional castable, verification by in situ analysis of castable plate was carried out. 25 mm thickness plates were set in a fireplace. The fireplace has been heated 4 hours per day during 100 days. Each day 20 kg of wood were being burned down. The humidity of the fire-wood was approximately 50 %.
The temperature of the castable plate inner surface during its exploitation reached up to ~800 °C, outer –
~500 °C. It has to be noted that the fireplace was working cyclically: the castable was being heated up from 20 °C up to 800 °C and cooled down again. It means that the influence of combustion products CO2, KOH, NaOH, H2O
upon the material might be different at the moments of fluctuation of temperature.
Fig. 1. Examples of castable products destruction in boiler (size (40 × 20 × 5) cm and (60 × 40 × 10) cm)
After 100 days of exploitation the plates were removed from the fireplace and analyzed in a laboratory. The analysis of castable plate showed, that the inner structure of the plate that was affected by combustion products differs greatly from the “cold” side, which leaned against metal shell of the fireplace. The castable of inner side of the plate was grey and fragile. The castable of the outer side was reddish and strong enough. It was determined that the hardness of the plate’s inner and outer sides differ greatly after the exploitation: outer – 12.5 HB while inner only ~6 HB.
The hardness of the castable fired at laboratory conditions depending on the temperature might reach up to 15 HB – 30 HB. As it might be noticed (Fig. 2) temperature negatively affects the hardness of the castable: after heating at 110 ºC – 800 ºC temperatures this value decreases 15 % – 45 % (it is due to the dehydration). It has
to be noted that the values of the castable’s hardness correlate very well with the values of castable’s cold crushing strength (Fig. 2). The difference between the values of exploited and not exploited castable hardness showed, that not only temperature but also other factors influence the material destruction during the exploitation. These factors we aimed to reveal in this work.
15 17 19 21 23 25 27 29 31 33 20 110 300 600 800 Temperature, oC Ha rd ne ss, HB 0 5 10 15 20 25 30 35 CCS , MP a hardness CCS
Fig. 2. The dependence of castable hardness and cold crushing strength (CCS) upon the temperature
According to the X-ray analysis of the inner and outer sides of exploited castable plate the following phases were identified: CA, C12A7, C3AH6 and C4AcH11 – calcium
carboaluminate hydrate (3CaO·Al2O3·CaCO3·11H2O). This
result shows that carbonation occurred in chamotte castable.
Aiming to understand the mechanism of the plate destruction the X-ray analysis of hydrated cement paste was carried out.
After one day of the hardening process of cement paste in the X-ray diffraction pattern of the specimen of composition A hydrates CAH10 and C2AH8 dominate,
ferrite phase as well as gehlenite (C2AS) are unchanged,
and mineral CA not identified.
After firing at the interval of 300 °C – 950 °C temperatures minerals C12A7 and CA were determined, and
C3AH6 was determined only after firing at 300 °C
tem-perature because it disintegrates at the temtem-perature of 360 °C (Fig. 3). The results might be explained by dehy-dration scheme of cement mineral hydrates [10]:
CAH10 → C2AH8+AH3 ; (4) C2AH8+AH3 → C3AH6+2AH3 ; (5) C3AH6+2AH3 → C12A7+СA+H↑ . (6)
Concluding the experiments it may be stated that in dehydrated paste C12A7 mineral forms. This mineral of
alumina cement is one of the fastest reacting with water [11]. The amount of this mineral depends on the temperature.
Amount of C12A7 increased when the specimens were
fired at the temperatures 600 °C – 900 °C and decreased after 950 °C. It can be notice that CA intensively crystallizes in the composition after firing at 900 °C.
The decrease of C12A7 after firing at 900 °C may be
explained by the intensive reaction of CA. The formula of the reaction is as follows [10]:
C12A7+5A → 12CA . (7)
When the specimens are fired and cured in humid conditions, C3AH6 considerably increases and C4AcH11
appears. It can be seen that C3AH6 persists up to 950 °C.
The biggest amount of C4AcH11 was identified in the
interval of 500 °C – 800 °C temperatures. The higher amount of C3AH6 and newly formed C4AcH11 in the
specimens, which were cured in humid conditions, explains the lower amount of C12A7 in them (Fig. 4).
0 25 50 75 100 125 150 175 200 225 250 275 300 300 400 500 600 800 900 950 Temperature, oC Peak height, r .u. C12A7 C3AH6 C4AcH11 CA
Fig. 3. Changes of peak height of minerals (C3AH6 (d = 0.229 nm), C12A7 (d = 0.220 nm), C4AcH11 (d = 0.76 nm) and CA (d = 0.297 nm)) representative peak after firing
0 20 40 60 80 100 120 140 300 400 500 600 800 900 950 Temperature, oC Peak height, r .u. C12A7 C3AH6 C4AcH11 CA
Fig. 4. Changes of peak height of minerals (C3AH6 (d = 0.229 nm), C12A7 (d = 0.220 nm), C4AcH11 (d = 0.76 nm) and CA (d = 0.297 nm)) representative peak after firing and curing in humid conditions
It can be concluded that C12A7 in humid conditions (as
one of the most active minerals of cement) easily hydrates and carbonates further forming C3AH6 and C4AcH11. The
volume of newly formed minerals is considerably higher than C12A7. It is obvious that cyclic changes (dehydration
at high temperatures, repeated hydration and carbonation in humid conditions, and dehydration of carbonates) of
C12A7 C3AH6 C4AcH11 CA C12A7 C3AH6 C4AcH11 CA
mineral volume may cause a destruction of castable’s structure during exploitation of fireplaces or furnaces.
The changes of minerals are different in composition B with the additive of microsilica.
After the hardening process of 24 hours C2AH8,
CAH10, unchanged gehlenite and ferrite phase as well as
CASmHn (it is the product of the reaction between
microsilica, minerals of cement and the water) were identified in the specimens of composition B.
After firing at the interval of 300 °C – 850 °C temperatures C12A7 and C2AS were observed. The
quantitative changes are reflected in Fig. 5. Unlike in hydrated cement paste (composition A) in fired specimens of composition B C3AH6 is not identified. C12A7 mineral in
composition with the additive of microsilica forms according to this scheme [10]:
CAH10+CASmHn+S → C2AH8+AH3+CASmHn+S; (8) C2AH8+CASmHn+S → C3AH6+2AH3+CASmHn+S; (9) C3AH6+2AH3+CASmHn+S → C12A7+CA+H+CASmHn+S. (10)
It has to be noted that an intensive additional crystalliza-tion of gehlenite is noticed at 850 ºC temperature. At this temperature the reactions take place according to this scheme [10]:
C12A7+S → C2AS+A . (11)
A small amount of CA2 and CA also appears (these
peaks are not shown in Fig. 5 and Fig. 6).
0 20 40 60 80 100 120 140 160 180 300 400 500 600 700 800 850 Temperature, oC Peak height, r .u. C12A7 C2AS
Fig. 5. Changes of peak height of minerals C12A7 (d = 0.220 nm) and C2AS (d = 0.285 nm) representative peak after firing
In the specimens cured in humid conditions (Fig. 6) mineral C3AH6 forms, but it’s amount is smaller than in the
specimens of composition A. After firing at 800 °C this mineral is not found any more in the specimens cured in humid conditions.
C4AcH11 mineral has not been identified in the fired
specimens, but a small amount of it was found in the specimens that were cured in humid conditions, however, after firing at 500 °C it is not observed (Fig. 6).
After firing at 850 °C C12A7 has not been identified as
well as amount of gehlenite considerably increased in any of the specimens (neither in fired nor in cured in humid conditions). 0 20 40 60 80 100 120 140 160 180 200 300 400 500 600 700 800 850 Temperature, oC Peak height, r .u. C12A7 C3AH6 C4AcH11 C2AS
Fig. 6. Changes of peak height of minerals (C3AH6 (d = 0.229 nm), C12A7 (d = 0.220 nm), C4AcH11 (d = 0.76 nm) and C2AS (d = 0.285 nm) representative peak after firing and curing in humid conditions
Mineral gehlenite at high temperature forms by the same scheme (11).
The analysis show that in the cement paste (composition A), cement hydrated phases carbonate up to the temperature of 900 ºC. As a result of the reaction between the additive of microsilica and the minerals of alumina cement „Gorkal 40“, which takes place during the hardening, drying and firing process, the resistance to moisture content as well as to the influence of CO2 is
increased: in the composition B that had been fired at 500 ºC – 900 ºC temperatures after the repeated hydration carbonates do not form any more and gehlenite, that crystallized additionally, grants better features of the composition due to its low reacting abilities.
CONCLUSIONS
1. The verification by in situ analysis of conventional alumina cement-bonded castable plates was carried out. It was determined that the compressive strength of the castable and its hardness is considerably lower in the area of the contact with combustion products than in the outer side.
2. Laboratory analysis of hydrated alumina cement paste showed that calcium carboaluminate hydrate (C4AcH11) forms in the paste, which was cured for two
days in humid conditions after it has been fired.
3. It was determined that the additive of microsilica increases the resistance of cement paste to the carbonation. In the specimens after firing and curing in humid condition small quantity of carbonated calcium hydroaluminate is identified only below 500 oC temperature.
REFERENCES
1. Majumdar S.D., Sarkar R., Vajifdar P.P., Narayanan S., Cursetji, R. M., Chatterjee, A. K. User Friendly High Refractory Calcium Aliuminate Cement Proceedings of the International Conference of Calcium Aluminate Cements Edinburg, Scotland, UK, 16 – 19 July 2001: pp. 467 – 476. C12A7 C2AS C12A7 C3AH6 C4AcH11 C2AS
2. Sakamoto, S., Kudo, E. Carbonation of Alumina Cement-Bonded castable Refractories Journal of Technical Association of Refractories (Japan) 23 2000: pp. 18 – 23. 3. Asano, S., Ishihara, S., Inoue, Y. Yougyou-Kyokai-shi 79
10 1971: pp. 357 – 364.
4. Blenkinsop, R. D., Currel, B. R., Midgley, H. G., Parsonage, J. R. The Carbonation of High Alumina Cement
Cem. Concr. Res. 15 2 1985: pp. 276 – 284.
5. Pérez, M., Vázquez, T., Triviño, F. Study of Stabilized Phases in High Alumina Cement Mortars Cem. Concr. Res. 13 1983: pp. 759 – 770.
6. Pérez, M, Vázquez, T. Study of the Strength Developed by Stable Carbonated Phases in High Glumina Cement Cem. Concr. Res. 14 1984: pp. 161 – 169.
7. Puertas, F., Fernandes-Carrasco, L., Blanco-Varela, M., Vázquez, T. Influence of KOH Solution on the Hydration and Carbonation of High Alumina Cement Mortars Journal of Materials Science 31 1996: pp. 2819 – 2827.
8. Garcia Alcocel, E., Garcės, P., Chinchon, S. General Study of Alkaline Hydrolysis in Calcium Aluminate Cement Mortars under a Broad Range of Experimental Conditions
Cem. Concr. Res. 30 2000: pp. 1689 – 1699.
9. Goñi, S., Andrade, C., Sagrera, J. L, Hernández, M. S., Alonso, C. A New Insight on Alkaline Hydrolysis of Calcium Aluminate Cement Concrete J. Mater. Res. 11 1996: pp. 1748.
10. George, C. M. Aspects of Calcium Aluminate Cement Hydration Proceedings of XIII Annual Symposium on Refractories March 25, 1994. Sheraton-West Port Inn St.Louis, Missouri: pp. 1 – 21.
11. Harr, K. W. Relationship between Mineral Phases C4AF and CA in Calcium Aluminate Cements with High Iron Content Proceeding of the International Conference on Calcium Aluminate Cements University Edinburg, Scotland, UK 2001: pp. 121 – 127.
Presented at the National Conference "Materials Engineering’2006" (Kaunas, Lithuania, November 17, 2006)